electricsheepafrica/ntd-schistosomiasis-filariasis

NTD Schistosomiasis & Filariasis Dataset

Abstract

This dataset provides 30,000 simulated NTD records (10,000 per scenario) from schistosomiasis and lymphatic filariasis endemic communities in sub-Saharan Africa. Each record contains 45+ variables including infection status, species, intensity, morbidity (hepatosplenomegaly, haematuria, hydrocele, lymphoedema), MDA coverage, treatment, and cure rates. Three settings: active MDA programme (19% schisto+), district hospital (38%), and endemic community without MDA (54%).

1. Introduction

Schistosomiasis affects 240 million people worldwide, with 90% in SSA. School-age children in hyperendemic areas have prevalence of 50–80% (Colley et al., Lancet 2014). Lymphatic filariasis causes hydrocele in 10–20% of endemic males and lymphoedema affecting 15 million people. MDA with praziquantel (schistosomiasis) and ivermectin+albendazole (LF) is the WHO-recommended strategy. Co-endemicity of schistosomiasis, LF, and STH is common across SSA.

This dataset is entirely simulated. It must not be used for clinical decision-making.

2. Methodology

2.1 Parameterization

Parameter	Value	Source
Schistosomiasis global burden	240M infected	WHO 2024
SSA share	90%	WHO 2024
LF global burden	51M infected	WHO 2024
Praziquantel cure rate	~75%	Colley et al., 2014
Hydrocele prevalence (endemic males)	10–20%	WHO 2024
MDA target coverage	>75%	WHO NTD roadmap
Co-endemicity (schisto + LF + STH)	Common	Hotez & Kamath 2009

2.2 Scenario Design

Scenario	Facility	MDA	Schisto+	LF+	STH+
MDA programme	Active MDA	75%	19%	4%	15%
District hospital	Limited MDA	30%	38%	10%	15%
Endemic community	No MDA	5%	54%	21%	39%

3. Schema

Column	Type	Description
id	int	Unique identifier
age_years	int	Patient age
sex	categorical	M / F
rural	binary	Rural residence
school_age_child	binary	Age 5–14
water_contact_frequency	categorical	daily / weekly / monthly / rarely
improved_sanitation	binary	Improved sanitation
occupation	categorical	farmer / fisher / student / domestic / other
barefoot_exposure	binary	Barefoot exposure
schistosomiasis_positive	binary	Schistosomiasis infected
schisto_species	categorical	s_mansoni / s_haematobium / mixed / none
schisto_intensity	categorical	light / moderate / heavy / none
lf_positive	binary	LF infected
lf_microfilaraemia	binary	LF microfilaraemia
sth_positive	binary	STH co-infection
any_ntd_infection	binary	Any NTD infection
haematuria	binary	Haematuria
hepatosplenomegaly	binary	Hepatosplenomegaly
portal_hypertension	binary	Portal hypertension
female_genital_schistosomiasis	binary	FGS
anaemia	binary	Anaemia
hydrocele	binary	Hydrocele
lymphoedema	binary	Lymphoedema
lymphoedema_stage	categorical	stage 1–4 / none
acute_adenolymphangitis	binary	ADL episode
mda_received	binary	MDA received
praziquantel_received	binary	Praziquantel
albendazole_received	binary	Albendazole
ivermectin_received	binary	Ivermectin
diagnosed_lab	binary	Lab-confirmed diagnosis
hydrocelectomy	binary	Hydrocelectomy surgery
mmdp_care	binary	Morbidity management
cure_schistosomiasis	binary	Schistosomiasis cured
cure_sth	binary	STH cured
reinfection_risk	categorical	low / moderate / high
disability	binary	Disability present

4. Validation

Key validation checks:

• Prevalence gradient: Schisto 19% → 38% → 54% ✓
• S. mansoni + S. haematobium: ~45%/40% split ✓
• MDA coverage: 75% → 30% → 5% ✓
• Water contact dose-response: Daily > weekly > monthly > rarely ✓
• Praziquantel cure: ~75% ✓
• LF morbidity: Hydrocele + lymphoedema in endemic adults ✓

5. Usage

from datasets import load_dataset
dataset = load_dataset("electricsheepafrica/ntd-schistosomiasis-filariasis", "district_hospital")
df = dataset["train"].to_pandas()

python generate_dataset.py --all-scenarios --n 10000 --seed 42

6. Limitations

• Simulated: Not derived from real MDA programme data.
• No geospatial: No GPS coordinates or transmission zone mapping.
• No egg counts: Intensity is categorical, not Kato-Katz egg counts.
• No onchocerciasis: Only schistosomiasis and LF modelled.
• No longitudinal: Snapshot, no multi-round MDA tracking.

7. References

1. WHO (2024). Schistosomiasis fact sheet.
2. WHO (2024). Lymphatic filariasis fact sheet.
3. Colley DG, et al. (2014). Schistosomiasis. Lancet, 383(9936):2253.
4. King CH (2010). Morbidity of schistosomiasis. Adv Parasitol, 73:1.
5. Hotez PJ, Kamath A (2009). NTDs in SSA. PLoS NTD, 3(8):e412.
6. WHO (2021). NTD roadmap 2021–2030.

Citation

@dataset{esa_ntd_2025,
  title={NTD Schistosomiasis and Filariasis Dataset},
  author={Electric Sheep Africa},
  year={2025},
  publisher={Hugging Face},
  url={https://huggingface.co/datasets/electricsheepafrica/ntd-schistosomiasis-filariasis}
}

License

CC-BY-4.0